Uses the Hugging Face pipeline to run predictions i.e text summarization on a block of text.
Details here: https://vishsubramanian.me/hugging-face-with-mlflow/
curl -X POST -H "Content-Type:application/json; format=pandas-split" --data '{"columns":["text"],"data":[["Sample Text"]]}' http://127.0.0.1:5000/invocations
or run predict_text.py